The System Agency AI Audit · Generated 11 Apr 2026
M&S · Agency AI Audit
38
/100
They know the language. They haven't done the work.
Brief: M&S asked for an assessment of their agency's AI pitch before briefing the CEO.
Six dimensions assessed
Strategic Framing
4/10
Presented a retail AI narrative that could apply to any FTSE 100 client. No reference to Stuart Machin's public AI commitments, M&S Food's demand forecasting work, or the Sparks loyalty programme.
Technical Credibility
6/10
Familiar with the tools. Can demonstrate a working prototype. The gap is deployment at scale — nothing in the pitch showed infrastructure for a 32-million-customer operation.
Data & Personalisation
3/10
Personalisation was mentioned twice. The Sparks loyalty dataset — one of the most valuable in UK retail — was never referenced. This is the most disqualifying gap.
Proof of Deployment
4/10
Referenced two case studies. Neither was verified in the room. Neither was in retail. Impressive work in other sectors doesn't translate without a bridge.
Team AI-Readiness
5/10
Senior team has the titles. Junior team is where AI literacy lives — and that layer wasn't present in the room or in the credentials.
Implementation Clarity
3/10
A three-phase roadmap was presented. No integration with the existing M&S digital stack, no indication of how outputs would connect to the buying or personalisation systems already in place.
Composite: 38/100
Commission a proof of concept before committing budget.
Ask for one deliverable — a working personalisation prototype on synthetic Sparks data. If they can't scope it, that's your answer.